Joint mobilization associated or not to other therapies reduces chronic musculoskeletal pain: a systematic review

BACKGROUND AND OBJECTIVES : Joint mobilization is a non-pharmacological technique used to treat chronic musculoskeletal pain. However, it is controversial due to a lack of studies comparing its effects on this painful condition. The objective of this study was to assess the risk of bias in clinical trials investigating the effect of joint mobilization on chronic musculoskeletal pain. CONTENTS : A systematic search on Pubmed, Cochrane Library, ScienceDirect, Scielo, PEDro, CINAHL, SPORTDiscus, LILACS, BVS, PsycINFO, Web of Science, and Scopus was performed on September 2019 from the combination of three keywords: Randomized controlled clinical trials that evaluated the use of joint mobilization associated or not to other therapies in chronic musculoskeletal pain treatment were included. Five thousand five hundred eighty-seven articles were screened, and 14 studies were analyzed, including 812 participants, with a mean age of 54 years, and female being the most affected. According to these articles, joint mobilization promoted the reduction of pain intensity in short and long terms, increase in range of motion, strength and function when used alone or in association with conventional physiotherapy. Regarding methodological quality, most of the studies were classified with low risk for selection, performance, detection and reporting bias. In the “other bias” item, which considered therapists experience o Estudo da Dor time and types of treatment applied, only one study presented low risk and other study presented an unclear risk. CONCLUSION : Joint mobilization seems to be an effective technique for the treatment of chronic musculoskeletal pain. Ho-wever, it is still necessary to investigate and compile studies with greater methodological quality, thus promoting greater


INTRODUCTION
Chronic musculoskeletal pain (CMP) is defined as a painful condition associated with dysfunction in muscles, ligaments, tendons, bones, and/or adjacent structures that lasts for more than three months 1,2 . Musculoskeletal conditions are the main cause of disability worldwide, with a prevalence ranging from 13.5 to 47% of the general population 3 . The main risk factors for this condition are advanced age, female sex, smoking, low schooling, sedentary lifestyle, poor social interaction, depression and anxiety 4 . This type of pain has a multifactorial etiology, which may have a definite cause (traumatic, ischemic, tumor, inflammatory, overload, overuse) or non-specific causes 5,6 . Several characteristics may be present in these patients, as generalized, diffuse and/or local muscle pain, physical and mental fatigue, a decrease of muscle strength, demotivation, sleep disorders, among others [6][7][8] , which may lead to increased health costs and reduced quality of life 9 . The treatment of CMP can be performed by a multiprofessional team and consists of pharmacological and/or non-pharmacological therapies 10 . The most used drugs are analgesics, non-steroidal anti-inflammatories, antidepressants, neuroleptics, anticonvulsants and myorelaxants 11 . Some non-pharmacological therapies involve physiotherapeutic techniques such as acupuncture, kinesiotherapy, electrotherapy, thermotherapy, phototherapy, spinal manipulation and massage therapy which aim to reduce pain and improve the quality of life of these patients 2,12 . Currently, these non-pharmacological treatments are being prioritized because of the lower risk of adverse effects. Manual therapy (MT) is one of these non-pharmacological techniques most widely used in the treatment of dysfunctions affecting the musculoskeletal system. It consists of a physiotherapeutic treatment that includes therapeutic massage, joint mobilization (JM), manipulation, among others 2 . JM is a technique used to treat musculoskeletal pain in the axial and appendicular skeletons. It is indicated to reduce pain, muscle spasms, reversible joint hypomobility, positioning/subluxation failure, progressive limitation, and functional immobility 13 . However, this technique has limitations and/or contraindications that consist of irreversible hypomobility, joint effusion and inflammation. The neurophysiological and mechanical effects caused by the use of joint mobilizations provide analgesia in patients with CMP 13,14 . Studies suggest 13,15,16 that JM activates the dorsal area of periaqueductal gray matter (PAG) of the brain, and this influences on pain perception. Studies show an immediate reduction of pain and an increase in sympathetic nervous system activity, suggesting an indirect relationship with the dorsal area of PAG and association between increased stimulation of sympathetic nervous system and reduction of mechanical pain threshold 13,[17][18][19] . An experimental study has shown an analgesic response accompanied by sympathetic nervous system activation after the electrical stimulation of the midbrain, originating in PAG 20 . Other clinical studies have shown a sympathetic-excitatory change combined with a hypoalgesic response after spinal JM, with increased skin conductance [21][22][23][24][25][26][27][28][29] , respiratory and heart rate 30 , and decreased skin temperature 31,32 . Changes in central sympathetic activity may be directly related to modulation response during therapeutic intervention 29 and it is also hypothesized that JM initiates the inhibitory mechanisms in the dorsal region of the PAG 13 . Using grade III JM in an experimental model of ankle joint inflammation, showed action through spinal blockade of serotonergic (5HT1) receptors, found in nucleus raphe magnus (NRM) and noradrenergic (alpha 2), located mainly in a small nucleus in the gray matter of the pons, the locus coeruleus. However, blockade of GABA or opioid receptors had no influence on the analgesic effect produced by joint mobilization 33 . These data may suggest that MA reduces CNS pain through non-opioid descending inhibitory pathways from the rostral ventromedial medulla and dorsolateral pontine tegmentum. More recently, the role of the nociceptive adenosinergic system has been shown to mediate the antihyperalgesic effect of MA by activating A1 adenosinergic receptors that predominantly mediate the effects of synaptic transmission in the superficial region of the dorsal horn 34 . There are several studies on the use of MT in diseases and other conditions, such as chronic spinal pain and osteoarthritis (OA) 10,13 . A recently published meta-analysis about manipulation and mobilization, specifically for the treatment of chronic low back pain, has shown that both therapies appear to be safe and that there is moderate quality in the studies that support the use of these techniques to reduce this type of pain 14 . Another meta-analysis addressed the use of manual therapy, exercise therapy (ET), or combined treatment for adults with cervicalgia. Quality of included studies was moderate and the authors concluded that combined treatment consisting of MT and ET does not appear to be more effective in reducing the intensity of resting neck pain, cervical spine disability, or quality of life improvement in adult patients with cervicalgia when compared to only ET 35 . Although it presents moderate scientific evidence according to the previously published studies, the use of JM in patients with CMP still presents controversies due to the lack of studies that directly compare its effects in this painful condition. Therefore, there is an even greater need for studies with methodological quality that is rigorous enough to indicate treatments in this area. Thus, in order to verify the existence of clinical trials related to this topic, this systematic review aimed to investigate and evaluate the effect of protocols for JM application associated or not to other therapies in the treatment of pain and motor performance in patients with CMP.

CONTENTS
As a PICO strategy, randomized trials with a control or placebo group that evaluated the use of JM associated or not with other therapies in CMP treatment were included in this review. The studies that presented participants older than 18 years old with chronic pain related to musculoskeletal dysfunction for time ≥3 months and who were treated with JM associated or not with other therapies were selected. Clinical trials comparing any type of JM with placebo or sham intervention, with no other type of treatment, mobilization as an isolated therapy or in combination with other conservative thera-pies have been included. Clinical trials comparing different protocols of JM (e.g., different degrees, series, repetitions, and/or body sites/segments) were also included. The exclusion criteria were studies with participants who presented oncological pain, headache, temporomandibular dysfunction (TMD), other painful conditions. It was also excluded studies with patients who were undergone to other modalities of therapies and/or mobilization under anesthesia or performed by machines as forms of treatment, use of JM only outside the site of pain, studies that did not report how long considered the pain condition as chronic and cross-over clinical trials. Studies that had no full-text accessible, and that was not possible to contact the authors, were also excluded. The primary outcomes evaluated were pain measured by a validated pain score scale, such as the visual analog scale (VAS) and numerical rating scale (NRS) and pressure pain threshold (PPT), measured by digital pressure algometer. As secondary outcomes were considered: the range of motion data (ROM) accessed through universal goniometer or inclinometer, muscle strength measured indirectly or directly through the isokinetic dynamometer and manual tests, functionality measured by validated functional tests, quality of life through validated questionnaires, such as SF-36 and QoL for general measures, adherence to treatment measured by the number of sessions that the individual performed and patient expectation /satisfaction measured through the patient's report and the Likert scale.

Protocol and register
This research protocol was registered in the International Registry of Systematic Reviews PROSPERO (CRD 42016046029). The inclusion criteria and analyses of studies were performed according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the Cochrane Handbook for Systematic Reviews of Interventions.

Search strategy
The studies were screened in the following electronic databases: Pubmed, Cochrane Library, ScienceDirect, Scielo, PEDro, CINAHL, SPORTDiscus, LILACS, BVS, PsycINFO, Web of Science, Scopus and Google Scholar. There were no restrictions on language or year of publication. The search was performed in September 2019 by combining the following descriptors: Manual Therapy/ Musculoskeletal Manipulations ("Musculoskeletal Manipulations"[mesh terms] OR (manipulations, musculoskeletal) OR (manual therapies) OR (manual therapy) OR (therapies, manual) OR (therapy, manual) OR (manipulation therapy) OR (manipulation therapies) OR (therapies, manipulation) OR (manipulative therapies) OR (manipulative therapy) OR (therapies, manipulative) OR (therapy, manipulative) OR (therapy, manipulation)), Chronic Pain ("Chronic Pain" Searches were remade immediately before the final analyses and additional studies were retrieved for inclusion. The reference lists of all primary studies were checked, and all articles were revised for additional references. Data collection and analyses were performed in accordance with the methods set out in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and the Cochrane Handbook for Systematic Reviews of Interventions.

Data selection and extraction
Two authors independently extracted data from all studies included using an Excel spreadsheet. This worksheet included extracting information about characteristics of studies, participants, methodological aspects, interventions, comparisons, primary and secondary outcomes, results, chronic pain condition, mobilization type, and association with other interventions. At this stage of the study, disagreements among reviewers were discussed with a third investigator.

Quality assessment
The risk of bias was assessed using the Cochrane Collaboration tool by two reviewers independently. Thus, six domains were evaluated: selection bias (random sequence generation and allocation concealment), performance bias (participants blinding), detection bias (evaluators blinding), attrition bias (results with incomplete data), reporting bias (selective reporting of outcomes) and other biases. For the judgment of this last type of bias (Other bias) the following aspects were considered: the number of physiotherapists who applied the protocols and their years of experience in the area (over one year), mobilization type (with different characteristics and/or combination with other therapies, presence of control group), no validated placebo for mobilization; compared with different therapies and with different application objectives. Each of these biases was classified as low risk, high risk, or unclear risk. Review Manager 5.3 was used for all quantitative analyses. The searches were remade immediately before the final analyses in September 2019 and additional studies were retrieved for inclusion in order to ensure the selection of the largest possible number of studies.

Included studies
The database search recovered 5587 potentially relevant references: Pubmed (258), PEDro (28), CINAHL (32), Cochrane Library (39), LILACS (0), Scielo (2), ScienceDirect (4412), Scopus (343), SPORTDiscus (19), Web of Science (145), PsycIN-FO (15), BVS (194) e Google Scholar (100). The search retrieved 5587 records of trials after removal duplicates, of which 30 articles were selected for full-text evaluation and 11 clinical trials met the inclusion criteria. Hand search on the reference lists of all primary studies was performed and further three clinical trials were selected; thus, 14 clinical trials were included for qualitative synthesis. Figure 1 shows the flowchart of the search and selection process in this review. Included clinical trials examining the JM intervention associated or not with other therapies in the treatment of CMP were publi-shed between 2008 and 2018. The total sample from 14 studies was 812 participants, ranging from 28 to 120 participants in each study. Of these, eleven (n=11) performed the sample calculation to estimate the number of subjects included [36][37][38][39][40][41][42][43][44][45][46] . The mean age of participants was approximately 54 years, ranging from 18 to 90 years. On average, a higher number of female participants were found (60,98%).
Some studies (25%) considered chronic pain after three months. Of these, two were about cervicalgias 41,42 and one about rotator cuff injuries 36 . Two studies (16,66%) had as inclusion criteria patients with pain for more than 10 years 44,47 . In another study (8,33%), chronic epicondylitis as of six months were considered for inclusion 15 . Table 1 summarizes the characteristics of the included studies. Other outcomes (n = 9) Other type of study (n = 4) Acute and subacute pain (n = 4) Mobilization and manipulation (n = 2) Included Studies included in qualitative synthesis and hand search (n = 14) Full-text articles assessed for eligibility (n = 30) Eligibility Selected records (Title and abstract) (n = 4001) Screening Records identified through database searching (n = 5587) Identification There was a significant difference in reduction of pain intensity in both groups Treatment with TENS and manual therapy produces a significant reduction in pain intensity, and there are no differences between these treatment groups

Pain intensity
The majority of included trials (11 of 14) measured pain intensity only at rest 15,36,[38][39][40][41][42][43][44][45]48 . One study measured the intensity of pain during movement (39), and one study evaluated pain intensity on movement and at rest 36 . Five studies evaluated the intensity of pain through the visual analog scale (VAS) 15 48 . In study 49 VAS was used to evaluate pain intensity associated with nociceptive flexion reflex test 49 .

Pressure pain threshold (PPT)
Some studies have evaluated PPT as a primary outcome 40,41,44,46,47,49 . This variable was measured through a digital pressure algometer at different points, according to pain location. The study 49 evaluated PPT in patients with chronic cervical disorders related to whiplash injury. Three measurements were conducted in the C6 spinous process at the median nerve trunk bilaterally elbow and the tibialis anterior muscles bilaterally at two times before and after therapy 49 .
The study 46 collected the measurements of PPT in the carpometacarpal joint at the bottom of the anatomical snuffbox and tubercle of the scaphoid bone. Three measurements were performed in the dominant hand of subjects with a one-minute interval, four times before treatment, 5 minutes after the treatment, 1 and 2 weeks after therapy. On study 44 the PPT was measured three times, with a rest interval of one minute between measurements, in bones connected to trapeziometacarpal joint, of scaphoid bone apophysis and hamato bone in four moments (baseline, immediately after treatment, one and two weeks after treatment) on symptomatic hand with OA carpometacarpal. Another study 47 consisted of a secondary analysis that focused on the contralateral hand (asymptomatic). The PPT was evaluated three times with a rest interval of one minute between measurements on the carpometacarpal joint of the contralateral thumb and the symptomatic hand with carpometacarpal OA, on scaphoid bone tubercle and hamato bone process in four moments (baseline, immediately after treatment, one and two weeks after treatment). On study 41 PPT measurements were performed three times, with an interval of 10 seconds between measurements in the following points: close to spinous process of cervical spine at medullary level treated on right side, with participant in pronated position, upper right trapezius muscle, between C7 and acromion, with sitting participant and trunk of the right median nerve at the  elbow, medial to the biceps tendon, with the elbow at approximately 30° flexion, with forearm resting on a support and participant sitting in three moments (before, after and follow-up).
The study 40 performed three PPT measurements with a 30-second interval between each application at the pain site in patients with plantar fasciitis. Algometry was measured twice (baseline and at the end of all care sessions). Five of six studies 40,41,[44][45][46] demonstrated a significant increase in short and/or long-term pressure pain threshold after JM application, associated or not with other therapies. Only one study 49 showed no significant difference after the use of this therapy.

Range of motion
Five of the included studies performed ROM measurement 38,[40][41][42]48 . Among studies that evaluated ROM with a universal goniometer, one article evaluated only active mobility 42 , and two other studies did not specified 38,48 . Active ROM was still evaluated through the inclinometer in a study 40 and through the cervical range of motion instrument 41 . Four articles showed a significant increase of ROM at cervical 42 , shoulder 48 , hip 38 , and foot 40 joints after JM application associated or not with other therapies. One study 41 did not find any significant difference for cervical ROM after therapy.

Muscle strength
Five studies evaluated muscle strength 15,36,44,[46][47][48] . In the study 36 , the isometric strength of the symptomatic shoulder was assessed for abduction, internal and external rotation through the Nicholas Manual Muscle Tester (Lafayette, EUA) performed with a dynamometer. After the demonstration and training test, participants were asked to push as much as possible against the dynamometer for 4 seconds while the evaluator provided a verbal stimulus 36 . The study 48 used muscle testing procedures by Daniels and Worthingham to assess the strength of the major shoulder muscles.
In this system, muscle strength is marked with a numerical rating ranging from zero, indicating no muscle activation, to 5 for the best possible response to manual resistance in a reduced range of the muscle group that performs the movement. Other studies 15,44,46,47 used the hand dynamometer to measure the strength of patients with chronic lateral epicondylitis and thumb metacarpal OA, respectively 15,44,46,47 . The studies from 44,46,47 also used the tweezer dynamometer to evaluate thumb strength 44,46,47 . Increased muscle strength after JM application associated or not to other therapies was observed in three studies 36,46,48 . Three other studies found no changes in strength after therapy 15,44,47 .

Functionality
Ten studies that were included investigated hip, shoulder, knee, and foot functionality. Nine of them used questionnaires 36,37,[39][40][41][42][43]48,49 and one used functional tests 38 , all with validation. The study 38 on the immediate effects of mobilization in patients with hip OA was the only that evaluated the functional indexes of patients through validated tests 38 . The Timed Up and Go test (TUG), which simulates some functional activities of daily living (from sitting to standing, walking and standing to sitting) 38 ; the 30s Chair Stand Test (CST), which assesses function and strength of the lower limbs 38 and the 40m Self Placed Walk Test (SPWT), which measures the time required to walk on short distances 38 . The other studies used the following questionnaires: Shoulder Pain and Disability Index (SPADI) to evaluate shoulder function in patients with chronic rotator cuff lesions 36 ; Knee Injury and Osteoarthritis Outcome Score (KOOS) to evaluate knee function/performance in activities of daily living (ADL) in patients with patellofemoral OA 39 ; the Neck Disability Index (NDI) was used to assess pain and disability of cervical spine in patients with chronic neck pain, cervicalgia and chronic whiplash injury disorders, respectively [41][42][43]49 ; the Upper Extremity Motor Activity Log Modified to measure upper limb functionality in patients with frozen shoulder 48 ; the Lower Extremity Functional Scale (LEFS) for assessing the functionality of lower limbs of patients with plantar fasciitis 40 ; the Oswestry Disability Index (ODI) that was used in patients with chronic low back pain to measure pain--related disability in the lumbar spine 37 and the Patient Rated Tennis Elbow Evaluation (PRTEE), which evaluated the functional outcomes related to pain and disability in patients with chronic lateral epicondylitis 15 . Improvement of functionality was observed in seven studies [36][37][38][40][41][42]48 after JM application associated or not with other therapies. There was no significant difference in functionality in three studies 39,43,49 after JM application.

Quality of life
Three studies assessed quality of life (AQoL) 36,39,43 , using different instruments, such as the Medical Outcomes Study 36-item short-form (SF-36) 36 , the Assessment of Quality of Life (AQoL) 36 , the ADL subscale of Knee injury and Osteoarthritis Outcome Score (KOOS) 39 and the 12-item health survey (SF-12) 43 . Only one study 36 showed a significant increase in the quality of life parameters after JM application. Two studies 39-43 did not show any significant difference in this outcome after therapy.

Adherence to treatment
Only two studies evaluated adherence of participants through medical records, considering the number of visits performed and the total number of visits that were pre-established in the study protocol 36,39 . Both studies showed good patient adherence to JM treatment, but no significant difference was observed between treatment and control groups.

Expectation/satisfaction
Three studies evaluated the expectations before the intervention, relating to their respective satisfactions after intervention 36,39,43 A five-point Likert scale was used (1 = much worse, 2 = slightly worse, 3 = no change, 4 = slightly better, 5 = much better) 36,39 , this data was registered in each patient's medical records. The study 43 evaluated the expectation before treatment from the concepts chosen by the patient: complete recovery, great improvement, partial relief or no expectation of relief. All three studies reported that most participants reported satisfaction and improvement with JM treatment performed.

Adverse events
Only three studies reported adverse events 36,39,43 . In the study 43 ,16.3% of patients treated with transcutaneous electrical nerve stimulation (TENS) (n=7) and 6.4% of those treated with manual therapy (n=3) reported treatment-related adverse effects. Three of them presented increased pain in the treated area and one showed the general physical condition of the group treated with TENS. Of those who received manual therapy, one patient reported clinical worsening during the first few days and the others did not specify symptoms 43 . Study 36 reported that during the intervention period, 17 participants out of 55 (31%) from the active group had adverse effects that included increased short-term pain during or after treatment (n=3), increase in short-term pain with home workouts (n=12) and slight irritation to tape used for postural taping (n=2). In the placebo group, five participants out of 61 (8%) reported adverse events involving increased short-term pain during or after treatment. During the follow-up period, 7 of 49 patients (14%) from the active group reported adverse events and included increased short-term pain with home exercise 36 .
In the study 39 adverse events were observed in seven participants who undergone exercise, education, manual therapy, taping intervention (skin reaction to the use of tape (n = 2)); edema after treatment (n = 2); pain in other areas after exercise (lumbar n =1; ankle n=1; another knee n=1). All adverse events were mild, with no need for medical intervention or treatment discontinuation (some bandage adjustments were performed and/or exercises were done by the physiotherapist). Thus, most of the adverse effects caused slight damages to the patients in included studies, were generally related to the increase of local pain immediately after the technique application, but without lasting for a long time.

Other variables
All studies recorded demographic data (sex, age, body mass index, height, among others). Other variables were also considered in some studies, such as thermal pain threshold 49 , nociceptive flexion reflex and pain related to this test 49 , pain catastrophizing 37 , level of anxiety, and depression 43,49 and use of drugs during treatment 39,42,43 . Two studies evaluated the levels of depression and anxiety in the participants through the Goldberg Depression and Anxiety Scale (GHQ-28). In one study, it was evidenced that 42.6% of the participants treated with manual therapy presented anxiety and depression 43 , and the other study showed that all individuals had high levels of anxiety and depression 49 .
The study 49 evaluated the thermal threshold, nociceptive flexion reflex and the pain associated with this test and showed that there was an increase in the nociceptive reflex flexion threshold in the group treated with JM and that there was no significant difference in pain during the reflex test nociceptive flexion and at the threshold of thermal pain. On sutdy 37 catastrophizing evaluation of patients with chronic low back pain was performed and pain catastrophizing interference was observed in the treatment of these patients Three studies reported the use of drugs during treatment 39,42,43 .
The study 39 evidenced similar use of analgesics, non-steroidal anti-inflammatory drugs (NSAIDs), glucosamines, and fish oil.
The study 42 use of pain medication was registered in five patients of both groups, and there was no significant difference between them. The study 43 details the periodicity of the consumption of NSAIDs, analgesics and muscle relaxants. In the group treated with manual therapy, 4.3% of patients took these drugs every day, while 12.8% reported weekly or monthly consumption.

Risk of bias
The studies were evaluated for the risk of bias (low, high, or unclear) in relation to six domains. Figure 2 summarizes the results of individual studies.
One study adequately described all domains and was considered as a low risk of bias 38 . One study presented a high risk for the selection bias by using an open randomization process (random list of numbers) 49 . Two studies presented an unclear risk for the performance bias because they did not present information about Bennell et al. 36 Beselga et al. 38 Crossley et al. 39 Farooq et al. 42 Horst et al. 48 Mayor et al. 43 Richer, Marchand and Descarreaux 15 Shashua et al. 40 Snodgrass et al. 41 Sterling et al. 49 Tavares et al. 37 Villafane et al. 45 Villafane et al. 44 Villafane et al. 45 the blindness of participants and researchers 40,49 . All studies had a low risk of reporting bias 15,36-44,46-49 . In the domain "other bias," 12 studies presented a high risk of bias 15,36,[39][40][41][42][43][44][45][46]48,49 . Six reported that more than one therapist performed the mobilization 15,36,39,40,43,48 , seven performed different types and techniques of JM without specifying the type of mobilization and/or combined with other therapies 15,36,39,40,42,43,48 , and 11 articles used different treatment controls, such as TENS, therapeutic ultrasound, laser, stretching or patient education 15,36,[39][40][41][42][43][44][47][48][49] . Three studies did not report the experience of the therapists who performed treatments 37,46,49 . One study was considered to be at an unclear risk because it did not report the number of therapists who performed the mobilization techniques and the time of clinical experience in the area 37 .
In general, the methodological quality of studies was considered a low risk of bias. The values referring to the percentage of articles classified in each degree of risk of bias are represented in figure 3.

DISCUSSION
According to the studies analyzed, JM seems to have better results in the management of chronic musculoskeletal conditions when it was used alone or in combination with conventional physiotherapy. From the 14 articles included, seven used JM interventions alone compared to a control group using simulated placebo therapy 37,38 , other interventions with appliances switched off or in non-therapeutic doses 41,44,46,47 , manual contact 49 or without treatment 37 and seven used JM in association with therapeutic resources used in conventional physiotherapy (TENS, TUS, strengthening, patient education and home exercises) compared to the same therapy without JM 15,36,39,40,42,43,48 . This shows the scarcity of clinical trials that use only JM as treatment, which makes it difficult to know the efficacy of this technique alone in various types of musculoskeletal disorders. In addition, the lack of validated placebo techniques for JM also difficult the discovery of new evidence about technique effect. Other bias 0% 25% 50% 75% 100% High risk of bias Low risk of bias Unclear risk of bias For primary outcomes (pain intensity and pressure pain threshold), JM promoted short-term 37,38,44,46,47 and long-term 41 pain reduction in different types of musculoskeletal pain when only mobilization was used as treatment. When applied in combination with other therapies in musculoskeletal lower limb dysfunctions (patellofemoral OA and plantar fasciitis) and spine (chronic neck pain chronic disorders related to whiplash injury and chronic low back pain), the studies showed a reduction of pain after treatment 37,39,40,42,43,49 . In upper limb dysfunctions (rotator cuff injury, frozen shoulder, lateral epicondylitis and thumb carpometacarpal OA), some studies have reported pain reduction 15,46 , and others did not observe any differences between groups 36,42,43,48 . Most of the articles also showed improvement of ROM, strength and function after JM treatment 15,[36][37][38][40][41][42]46 . Based on these studies, most of them with good methodological quality, JM can be considered an effective therapeutic resource in reducing pain and improving the functionality of patients with CMP. There is insufficient evidence to determine the effect of JM on quality of life, adherence and patient expectancy in the treatment of CMP. Previous reviews and meta-analyses show chronic musculoskeletal conditions studied separately and involving manual therapy as a set or combination of manual techniques (neck pain 35,50 , lower back pain 14,51 , patellofemoral pain 52 , impingement syndrome 53 , hip OA 1 , or using a particular technique of JM for pain treatment 54 . In addition, it presents restrictions on language and publication period of selected articles, and the search was performed in a small number of databases, which makes these studies limited. Differently, the present review performed the search on 12 databases, without year and publication language restrictions, which reduces the risk of loss of some article at screening and, consequently, provides greater access to the data extracted from studies already published about this theme. The prevalence of common musculoskeletal conditions has a strong relationship with age, being directly proportional to aging, with an increase of painful complaints due to the wear of the musculoskeletal system 3 . In this review, neck pain was more prevalent among included studies [41][42][43]40 , in contrast to the number of back pain studies in the literature, who had only one study included 37 . This may have been caused by not meeting the inclusion criteria in this review or the fact that manipulation seems to be the most appropriate technique for low back pain according to current revisions 14 . Later appear the injuries that affect the shoulders, such as rotator cuff disease 36 and frozen shoulder 48 and to a less extent, diseases such as lateral epicondylagia 15 and plantar fasciitis 40 . OA is a common condition shown in studies in different body regions and patellofemoral 17 , carpometacarpal [44][45][46] and hip 16 . Autors 51 in a systematic review (SR) about JM and exercise efficacy for different stages of non-specific low back pain, found that JM (being the high speed and low amplitude manipulation of the pelvic loin region that presents evidence of moderate support compared to mobilization and soft tissue techniques including "myofascial," "miotensive" or "harmonic" techniques in this same body region), in combination with specific and/or general exercises, or usual medical care, are better than any of these isolated interventions.
In agreement, study 14 in a meta-analysis on manipulation and mobilization in the treatment of chronic low back pain, have observed that the manipulation of high speed and low amplitude (thrust) is the most recommended. However, the search conducted in this SR showed year restriction (January 2000 to April 2013), language (only articles in English), and a few electronic databases (Medline, Cochrane Register of Controlled Trials, PEDro, CINAHL, PsycINFO, and ICL). In this meta-analysis, a specific population was not defined since there was no homogeneity in the causes of low back pain, which implies a greater heterogeneity of studies and consequently a higher methodological bias and less applicability of these data in clinical practice.
In an SR about the use of exercise in mechanical neck pain, study 55 concluded that there was no evidence of high quality, indicating that there is still uncertainty about the efficacy of specific strengthening and resistance exercises for neck pain. Study 50 , in SR about JM efficacy and exercise for nonspecific cervical pain, emphasized the importance of performing combined treatment (JM plus exercises), being better compared to JM or exercises alone. The authors also showed that JM does not need to be applied at the symptomatic level to improve pain and can then be applied at adjacent levels according to the irritability status 50 . These reviews also show that manipulation is less indicated in cervical disorders, recommending that the thoracic manipulation or mobilization should be performed. This agrees with results found in studies included in this SR [41][42][43]49 , who considered mobilization as a treatment with good results in patients with nonspecific mechanical neck pain and with whiplash injury-related cervical disorders, applied either alone or in association with conventional therapy.
In studies about chronic pain, pain intensity is considered the primary outcome in most clinical trials 56 and should be investigated both at rest and during movement. From the included studies, 12 evaluated only pain at rest 15,37,38,[40][41][42][43][44][45][46]49 , one study assessed only pain in motion 39 , and only one has evaluated pain at rest and during movement 36 . Both the intensity of pain in rest and during movement must be evaluated because there are pain conditions that do not occur or have lower intensity at rest, which may mask the evaluation and generate a bias in the study.
Therefore, it is very important to perform pain assessment in these two conditions in order to have more reliable data about the pain of a population. In the reviews cited above, positive results were also observed for the reduction of pain associated with the use of manual therapy/ mobilization and combinations with exercises 50 . In a previous SR about the use of physiotherapeutic treatment in subacromial pain, it has been shown that exercise therapy should be the best treatment for pain reduction and improvement of function and range of motion. However, the addition of joint mobilizations to exercise can accelerate the reduction of short-term pain 53 . The same can be observed in studies included in this SR that combine the use of JM with conventional physiotherapy 36,39,30,43,43,48 . People with chronic pain present reduced functionality, and this loss can be evaluated through functional index instruments 18 .
Of the 14 included studies, only seven assessed functionality [36][37][38][39][40]48,49 , which shows a deficiency in studies regarding the evaluation of this variable. This is evidenced by the Initiative on Methods, Measurement and Evaluation of Pain in Clinical Trials (IMMPACT) and Outcome Measures in Rheumatology (OME-RACT), which bring the importance of assessing both pain and functionality of patients 17 .
A systematic review investigated the risks of manual treatment on the vertebral segment and concluded that serious adverse events are rare, the most common are mild, and these are associated with a greater amount of spinal manipulation 57 . Although some studies included in this SR 36,39,43 have investigated adverse events as a secondary outcome and have reported as minimum, there are not enough data to prove this variable due to heterogeneity of the studies. This review presented some strength points, including the development of the question and the population table, intervention, control/comparison and result (PICO strategy), use of a systematic, explicit and transparent methodology, incorporating internal validity evaluation (risk of bias), independent methodological evaluation by a third reviewer for each of the technical steps involved in the review phases, and a comprehensive survey in most databases, without restrictions. None of the authors reported any conflict of interest. Nevertheless, there are some limitations in this systematic review that make it impossible to conduct a meta-analysis. Although most of the included studies present low risk of bias, there was a significant heterogeneity regarding the protocol of JM application (different degrees of mobilization, series, repetitions, body segments, type of mobilization -Mulligan, Maitland, Kaltenborn and passive), duration of the chronic condition, treatment performed by more than one professional with different experiences and training times, compared to control groups or placebo using other techniques or treatments that are not related to manual therapy and outcomes assessed in different ways. From that, future clinical trials should aim to use only the JM technique in experimental groups to standardize the application protocols of the techniques and validate placebo techniques with the use of manual therapy. Based on that, studies can be performed with greater methodological accuracy regarding the application of the technique and can generate greater and more reliable results.